December 25, 2012
Merry Christmas, for those who celebrate it. I don't really...
December 22, 2012
I pulled the disk and replaced it after mucking with the Molex power connectors
a bit, and badblocks the disk... it turned out fine! Weird. I reintroduced
the disk back into the array and made the 250G disk into a bona fide hot
spare instead of the originally manual-insert hot spare. The disk still seems
to be holding up...
December 21, 2012
It's not the end of the world as we know it... but one of my hard drives
in Doujima isn't feeling fine... I hotspared the 250G to take over for the
failed 2M cache 120GB Maxtor. However I don't really think it's dead...
December 8, 2012
Ugh... While trying to build a better power supply that could fast charge
my Nokia 5230 (Nuron), I blew the phone's charging circuit somehow. I can no
longer charge my phone through the normal charging port, and it appears I can't
power the phone either :( Time to buy a new phone. *sigh*
I thought this phone was fine. A bit slow but allright. Now I wonder what
phone I should get. At least I have no contract to worry about, which was
the whole point of going T-Mobile.
November 17, 2012
On a note from badcaps.net while trying to
also fix the dead Antec SP400, I was able to determine the controller chip
was also dead along with the capacitors! Replacing it restored operation
to the PSU!
The Antec SP400, unlike the SP300, uses a KA7500C SMPS chip (which was replaced
with a working DBL494). It also has a TPS3510 supervisor like the SP300.
I believe the mode of failure of the PSU was the +5VSB
circuit's capacitors dried out, causing a massive voltage spike to pass, frying
the KA7500C.
November 15, 2012
Bingo. Repaired the Antec SP300. For the heck of it I oiled the fan and it
freed up and could spin again. Analyzing the SP300,
the PSU uses a Unitrode UC3842 (or was it a UC3844?) based controller and
a TPS3510 PSU supervisor which handles power on and overvoltage (shutdown)
and undervoltage (shutdown, assuming a short, as well as deasserting Power
Good) protection. It also has a temperature sensitive fan that senses
the temperature near the low voltage Schottky rectifiers' heatsink.
After finding and replacing a boat load of bad caps (about 70%
were bad) it still did not work. After the caps were replaced, at least
the +5V Standby was restored. However I was able to make it "blip" on a few
rails and the UC3842 was getting power... for a split second. It was shutting
down! Monitoring the TPS3510 I found that the +12 rail was not going high,
which would cause a shutdown. I found that the +12 line was 0.3Ω to GND!
After pulling out lots of parts I found that the toroidal filter
inductor was shorted to +5V amongst other things, I had another one from
another PSU that I parted and put it in... and replaced one last 1000µF
16V capacitor that should be low ESR which read 4 ohms...
After all that, the PSU sprung back to life! It powered a 50W headlight just
fine on its 12V rail.
I think I will continue to repair the SP300 as needed, it has a fairly simple
and common controller circuit and its output voltage protection makes it
safe when it fails.
November 7, 2012
Thought about re-capping the Antecs. Now I have both an SP300 and an SP400
that have failed. Both appear to have good transistors still so perhaps they
can be repaired by capacitor replacement. I wonder if the Antec Smartblue 350
I have will ultimately suffer the same fate. At least that the transistors
didn't fry...
November 5, 2012
Sorry about downtime of Doujima, PSU failed... I had to restart it a few hours
earlier to clean up some stuck processes. Fan was seized but I ignored it,
thought it was working when it turned quiet... WRONG. The poor Antec Smartpower
300W cooked. Not sure if I want to repair it - it was a fairly good PSU but
damage looks extensive; something let out the magic smoke, I could smell
it... Using a temporary PSU solution for now.
November 1, 2012
Started building a second ESR meter using the meter I snatched from the hamfest.
October 30, 2012
Returned from CA. Went to Jameco the day before, not much of a retail store...
Restocked on a bunch of capacitors.
October 23, 2012
Returned from MD.
October 20, 2012
Went to the Calvert County Radio League Hamfest.
Went with my friend from MD. Got a few capacitors, a two ethernet cards,
a junk meter, and some other misc stuff. Going to use the meter with another
ESR meter I'm building.
October 4, 2012
Bought new tires for my Subaru. Set of four Michelin Primacy MXV4 all-seasons
after about 60K miles on the Goodyear Assurance Tripletreds.
October 1, 2012
Finally designed and assembled an ESR meter. Cut a hole into a plastic case
and stuck an analog meter into it. Etched a PCB for it.
September 19, 2012
I donated Rinoa, the Gateway Solo 5350, to Goodwill. It's been a long time but
it was time to go. I just hope that Hinoki will do the job, starting to run out
of machines with legacy serial and parallel ports. Rinoa was the only laptop I
had left with a normal floppy drive. I did end up getting a USB floppy but
that's just not the same thing. Fujin's floppy drive isn't completely normal
either. Rinoa will be missed. The name will be recycled I suspect,
will be a few months...
After moving Mikuru around, I think that Chii still has its advantages.
Definitely smaller in the XY dimensions though Mikuru is thinner (Z dimension).
Just can't really run WoW on Chii, though it works fine on Mikuru.
September 2, 2012
Got sick of one of my LCDs that have been flaking out, it was whining when off
and frequently came up with a garbled screen. I opened it up again
found two capacitors that were leaking pretty badly (not leaking fluid, but
leaking charge). I replaced one with a similar one (1000µF 10V with
1000µF 16V) and the other I was kind of stuck, replaced the 470µF
25V unit with two 220µF 25V units in parallel using pads that were
available on the board... Not exactly the same but good enough. The monitor
seems to work just fine once more.
I also noticed that Windows wasn't even using the resume partition that the restore CD created on Mikuru. I zapped it and made it into a Linux swap/resume partition instead. Now I know that the 32G SSD on the machine is just sitting there completely fallow...
August 26, 2012
Installed SSD and restored Windows to the SSD.
Somehow Windows decided to put the hibernate partition on the SSD which isn't
exactly what I wanted... need to better distribute this...
August 24, 2012
Here we go. Disabled Intel Smart Response on Mikuru to decouple the 32G SSD
and the 500G HDD. Still trying to figure out how to get the hibernation file
off the 32G SSD.
Probably will partition:
256MB or so - Grub Boot/EFI partition (will need to doublecheck)
20GB Linux
20GB Windows 7
Remainder - ?? Need a filesystem that works with both Win7 and Linux...
August 23, 2012
After a bit more than a week of waiting I got the Crucial M4 SSD, and it
indeed is a 7mm thin drive. I also got a USB3 external enclosure.
I got 183MiB/sec through SSD->SATA->USB3, and
34MiB/sec through USB2. Now the plan is to install this in Mikuru, the Envy4.
July 20, 2012
I broke down and bought a HP Envy4t-1030US Ultrabook to replace my netbook
and notebook. Now I don't have an excuse to keep several laptops...
(Rinoa, Rukia, and Chii).
Hinoki still fits a space the Envy doesn't fit... The Envy does not have a
IEEE1284 parallel or RS232 port... This machine is thin, not quite as thin as
others I've seen but nevertheless it's thin. Unfortunately can't speak much
of its speed other than it seems slightly slower than my work laptop, though
it's expected - the work laptop has a discrete GPU as well. Then again this
machine has a Intel HD4000 which is not too bad.
One thing that seemed a little confusing is that the hard drive on it already
has all four partitions used - which means it's going to require a bit of work
to get Linux installed. I'm still not sure why all four are used - two are
clear: one's for recovery and one's for the main install. However the other
two I'm not so sure about.
Ultrabook is named Mikuru, from the anime "The Melancholy of Haruhi Suzumiya."
July 8, 2012
These past two days have been innundated with torrential rain. No hurricane or
tropical storm, but a lot of moisture from the skies fell. It's a nice soaking rain.
July 7, 2012
Found a Logitech Quickcam Express. This is a really, really crappy USB webcam.
352x288 using an Agilent sensor, manual focus, USB1,
probably around 10fps or so. The picture quality
is really poor. Anyway it appears that originally this camera used
quickcam.ko as the driver.
I got quite panicked that this driver is very
much abandonware. However I found that it is now is part of the
gspca driver - specifically, gspca_stv06xx - which is part of the 3.2 kernel
at least. After using this module this works just fine under
Linux/GNOME/Cheese.
July 5, 2012
I got the new PLCC extractor ... it was a Made in China like most of these
things most likely, for a lot less than the $10 RatShack wants.
July 4, 2012
Happy Independence Day...
July 3, 2012
I got the P2103NVG mosfet transistors. That was fast from China...
June 30, 2012
After tremendous futzing with options, discoverred why things were so slow.
Found that only static IP machines on the /29 were slow (including Doujima
which was a big problem.) Discoverred if I disabled NAT on the GT701WG it
made it run much better. This has the unintended consequence of making
the wifi more or less useless on the Actiontec (well, it can be used
for the static IPs...)
Unfortunately I was hoping I could use the IP address eaten by the gateway
by NATting through it. I guess not. Well, I guess the WRT54G has to do the
NAT routing. And I also found out that Doujima's NAT is not working properly
either, so that needs to be fixed...
Was also trying to set up virtual SSIDs on the WRT54G. It didn't seem to like
more than 2.
June 29, 2012
Was able to actually get 5Mbit DSL down. Finally, faster downloads...
June 26, 2012
I orderred a second 3Mbit DSL link because downloads were too slow.
The Centurylink guys came and installed a second DSL link to augment the
slow 1.5 Mbit download link. However my 1.5Mbit link started to really
act up. Not sure why it decided to puke out... The tech mentioned that
I could get 5Mbit down, it was close but possible. All hopes this will
happen, having dual links is going to be hard to manage.
June 24, 2012
Ended up orderring another PLCC extractor and a few MOSFETs through eBay.
Now, whether to get a DSO or not...
June 23, 2012
One of the UPS battery packs died again. Need to get a new battery pack I
think. The Werker battery I bought seems still to be fine despite being 2 years
old at this point. Maybe Powersonic batteries are crap? I don't know...
June 9, 2012
Fire! High Park goes up in flames...
May 2, 2012
Ordered a used laptop m/b from Ebay.
April 17, 2012
Decided to hack up a variable PSU based on an MC34063A. Used an inductor from
an old PSU. Didn't realize that the max current of 1.5A of the SMPS IC was
unfortunately halved because of the max 50% duty cycle of the switch...
April 15, 2012
Happy tax day...
April 13, 2012
Found a Cigarette lighter to USB adaptor from Dollar Tree. I was thinking
it was a cheap 7805-based charger but it actually had a MC34063A switch
in it! That would up the efficiency quite a bit.
March 7, 2012
Went to the Longmont Amateur Radeo Club hamfest. At first it looked like a
pretty uneventful, but eventually found some treasures. Got a a bag of
variable resistors/potentiometer and knobs for them...
I was really short on knobs.
I found some 4.096V voltage reference IC's
that I'll see if I can use to calibrate my multimeters.
Unfortunately some of these were surface mount units and cause a bit of trouble
wiring to them, but oh well.
Grabbed a few MC34063A SMPS IC's to make small PSUs (their 40KHz switch speed
is pretty slow though).
The biggest find was getting a few Rail-to-Rail I/O op amps (lm6142
and tlv2374) ! These
patent encumbered op amps take inputs from/to both rails and outputs to both
rails before flatlining. Much like the classic LM358 except it can deal with
voltages close to both rails instead of just GND.
March 6, 2012
After a bit of anticipation I went to go see the Denver Nuggets play the
Phoenix Suns at the Pepsi Center. Well, it wasn't the game, more of time with
the nicest person in the world... and probably cutest too...
February 18, 2012
Well, was using Ouka today and found out that the keyboard seemed a little
flaky. Turned out the two AAA batteries in it were dead. Replaced them
with Duracell expire in 2008 cells.
They were Energizer expire in 2008 cells...
February 12, 2012
Went to the Aurora Repeater Assosciation hamfest. Not way too much neat there
but got more 50K pots (think from the same guy who I bought from last time I
went to ARAfest), some more 47µF 50V capacitors (ran out..).
I also got some DIACs for a quarter, never played with them before,
but might be interesting to play with TRIACs...
One really odd buy that I wasn't sure of was a bag of SIMMs and DIMMs.
Seemed to get in the grab bag:
3 4MB 72-pin SIMM (worthless, but could stick in printer?)
1 32MB PC?? DIMM (utter junk?)
1 256MB PC133 DIMM
2 512MB PC133 DIMM (interesting...)
1 256MB PC133 SODIMM (what I was most interested in, for my Geode or GW)
1 512MB PC133 SODIMM (bad?)
1 256MB PC266 DDR SODIMM (tested works, but not much use)
2 1GB PC5300 DDR2 SODIMM (haven't tested)
I thought there weren't any DDR2 SODIMMs, but 1G DDR2 SODIMMs aren't very
useful anyway. Thought they were all DDR1 but the DDR1 notch is eversoslightly
off compared to a DDR2 notch and hard to discern by eyeballing. One thing I
should have guessed with is that there were lots of BGA chips on the board,
and many DDR chips are TSOP. Now I have three 1GB DDR2 SODIMMs unused (one
when I displaced the netbook's 1GB DIMM with a 2GB unit.
I stuck the 256 and 512 PC133 DIMMs into my Via board (which
seems to be the only one that will work with these large DIMMs) for a total
of 1.25GB RAM, ready to compile Firefox...
but it will still take forever with a 500MHz CPU and a slow @$$ VIA chipset...
I stuck the 32M DIMM into the Celeron 1364... not much of a boost but at least
it's something - from 384MB to 416MB.
Next Hamfest is in April, LARCfest. Though BARCfest is held in the same place,
they tend to have more discarded items there...
January 21, 2012
Went to Northern Colorado Area Repeater Club Hamfest.
Nothing great there, but did get a free lite-on cdrw drive that was stuck
(same problem as another lite-on drive I have), a free dead Asus motherboard
with Athlon XP3200+ (not sure if it's board or cpu or both that are bad),
a free 128MB SD card (New), two old 12V 7AH AGM batteries, a bag of 20 pin
IC sockets (for AVR90S2313) and two mystery PCBs I was planning to part.
After studying them they turned out to be a UPS board and a garage
door opener board.
January 10, 2012
Checked the displaced UPS battery, it looks like one of the cells is now
defunct and will only generate 4V. Time to get a replacement...
January 3, 2012
Woke up to a beeping UPS. Battery failure. Found one of the three batteries
(a Powersonic) failed, seems like another of the cells dried out.
Added water and trying to recover it... Fortunately the battery dried out when
charged - which means the plates should be lead and lead dioxide - and can't
sulfate. At least I hope that's the case. It would mean that the remaining
acid however would be quite concentrated, that may be an issue.
Placed spare battery that dried out before back in its place and pressed the
button for 3 seconds, it silenced the alarm and retested the set.
Been playing with the freshly watered pack, it seems to be holding up.
Definitely needed quite a bit of water to fill one of the cells,
but the other two cells were quite full already - seemingly some acid bubbled
out during overcharge.
I also inadvertantly unplugged the serial cable to the UPS somehow and no
wonder Doujima did not notice the battery failed. Upsd didn't behave the way
I expected with the cable unplugged, will need to fix that.
I have one Werker and three PowerSonic batteries for the UPS,
three of which are used and one spare.
January 1, 2012
Happy New Year! Well this is it, 2012... The world is (NOT) going to end.
December 26, 2011
Was looking at the nonfunctional circline fluorescent lamp with
electronic ballast again. Using a DMM and my PSU (to validate my results),
found that both output MOSFETs (ST P7NK40Z lot W18410, made in China)
and a current sense resistor (brown orange gold, assumed, so that would be
1.3 ohms...) were all fried. I'm not sure about the IC (UBA2021P),
it's very likely that it too is dead.
Not sure what to do with it now, it's probably trash time
since unfortunately getting new MOSFETs at $1 or so a piece
does not seem like a good idea. This was a 32W
bulb and actually had a controller IC instead of a transformer
feedback oscillator. The two MOSFETs actually had built-in zener protection
and were gate protected, so not exactly cheapo. Perhaps I'll
try putting in the thought-defective and used units to see if they will work.
Also changed my garage door remote control battery once more. Not sure if the
battery was really dead or not, but the transmitter just didn't have
much range. The battery measured 3.1V open, 2.7V with 560 ohm resistor load.
The new cell was 3.4V open, 3.2V with 560 ohm load.